MicroRNAs play an important role in the regulation of gene expression by binding mRNA targets causing their degradation or blocking their translation. Several genes has been found to be implicated as miRNA targets in different types of malignant tumors suggesting their involvement in cancer pathogenesis. Detecting direct miRNA-targets associations is not straightforward as in principle targets expressions are not altered except when they are completely repressed by the degradation complex. In this paper we propose an approach to identify direct miRNA-targets associations hypotheses by means of indirect association measures such as mutual information. Indirect regulons of miRNA and Transcription Factors (TFs) are compared with the Fisher's exact test to identify potential co-regulations which may constitute potential miRNA-TF direct associations. We apply the method on two cancer datasets, Colon and Lung, drawn from the Cancer Genome Atlas (TGCA) obtaining promising results.